Short-Long-Term Propagation-Based Video Inpainting

Shibo Li, Shuyuan Zhu, Yuzhou Huang, Shuaicheng Liu, Bing Zeng, Muhammad Ali Imran, Qammer H. Abbasi, Jonathan M. Cooper

Abstract: In this article, we propose a new method to inpaint videos with removed regions. Our method was developed based on combining both short-term propagation-based inpainting (STPI) and long-term propagation-based inpainting (LTPI) modules. The STPI module is designed to in-fill an image from a single frame with local reference information, while the LTPI module uses multiple STPI modules to inpaint the whole video, giving high temporal consistency and low complexity. With both of the proposed modules, the correlated spatiotemporal information of frames can be propagated throughout the video, offering reliable short-long-term source information for inpainting. The experimental results demonstrate that our proposed method provides better results when compared with the state of the art.